220 



ON THE PERIODICITY OF THE AURORA BOREALIS. 



I translate the following passage from Boue : — 



"If the monthly frequency of northern lights and earthquakes is compared for a series of years, a decided and gen- 

 eral correspondence is discovered between the tables of northern lights given by Mairan, Bertbolon, and Muncke, and 

 those which Perrey has made of earthquakes since the fourth century in northerly Europe and Asia, in western Europe, 

 as well as in eight different basins and in south-eastern Europe. I do not mean a relative coincidence of numbers, but a 

 somewhat similar mutual proportion of numbers for both classes of phenomena in each month of this series of years. If 

 the tables of Perrey are compared with Hansteen's registers of the northern lights, the greatest disagreement falls upon 

 the months of February and March as well as September and October. But May, June, July, and even August retain 

 the same character for both classes of phenomena, as the months of minimum frequency. Decidedly in July and pre- 

 eminently in June both are of great rarity." ' 



Baumhauer 2 has classified the recurrences of meteoric stones and fire-balls as follows : 



1 Sitzungsberichte der Kaiserl. Akad. der Wissensch. Wien. XXII. p. 434. 

 2 Pogg. Annalen der Physik und Chemie, LXVI. 477. 



